neighbayes.models.SARLogit

class neighbayes.models.SARLogit(formula=None, data=None, y=None, X=None, W=None, *, unit_effects=None, time_effects=False, dynamic=False, latent=None, unit_col=None, time_col=None, N=None, T=None, priors=None, logdet_method=None, robust=False, durbin_vars=None, logdet_decomp='auto', logdet_adapt=True)[source]

Bayesian SAR logit: the latent SAR, or the reduced form with latent=False.

Parameters:
formula : str, optional

Wilkinson-style formula, e.g. "y ~ x1 + x2". Requires data. An intercept is included by default; suppress with "y ~ x - 1".

data : pandas.DataFrame or geopandas.GeoDataFrame, optional

Data source for formula mode.

y : array-like, optional

Binary dependent variable of shape (n,). Required in matrix mode. Must contain only 0 and 1.

X : array-like or pandas.DataFrame, optional

Design matrix. Required in matrix mode.

W : libpysal.graph.Graph or scipy.sparse matrix

Spatial weights of shape (n, n).

priors : dict or SARLogitPriors, optional

Override default priors. Supported keys:

  • rho_lower (float, default -0.999): Lower bound of the Uniform prior on \(\rho\).

  • rho_upper (float, default 0.999): Upper bound of the Uniform prior on \(\rho\).

  • beta_mu, beta_sigma (float or array, default Gelman et al. 2008): Normal prior on \(\beta\), on the logit scale (intercept at logit(mean(y)), scale 2.5; slopes 2.5 / sd(x_j)).

logdet_method : {"eigenvalue", "cheb", "aaa", "spline", "stochastic"}, optional

How log|I - rho*W| is computed over rho: summed over the eigenvalues of W; interpolated (Chebyshev, AAA rational, or cubic spline) through values from a sparse factorization, exact to the interpolant’s tolerance; or expanded stochastically with no factorization, the one method with sampling error. None (default) selects from the size and symmetry of W: "eigenvalue" for n <= 500; "aaa" up to n = 60000 (NEIGHBAYES_LOGDET_FACTOR_MAX_N) unless the factorization’s fill-in would dominate; "stochastic" beyond.

logdet_decomp : {"auto", "chol", "lu"}, default "auto"

The factorization an interpolating method uses: sparse Cholesky (undirected W; scikit-sparse CHOLMOD) or sparse LU (KLU or UMFPACK, chosen by the sparsity pattern). "auto" takes "chol" where the method has it and W is undirected, else "lu". Ignored by "eigenvalue" and "stochastic"; "spline" on "chol" raises.

robust : bool, default False

Not supported. Raises NotImplementedError if True.

latent : bool, default True

The latent SAR; False fits the reduced form. The latent SAR is

\[y_i = \mathbb 1\{y^*_i > 0\}, \qquad (I - \rho W) y^* = X\beta + \varepsilon, \qquad \varepsilon_i \sim \mathrm{Logistic}(0, 1),\]

whose ρ = 0 member is the logit. It is sampled by an exact Gibbs sweep over the latent utilities and the Kolmogorov scales of the logistic (Albert & Chib 1993; Holmes & Held 2006), which needs augax: on JAX by default, the utilities by colour class, or with backend="numpy" one at a time. Its impacts are on the latent index, scale="logodds".

Notes

The reduced form parameterizes the log-odds as the deterministic mean-propagator eta = (I - rho * W)^{-1} X @ beta (no latent noise field), and augments the logistic likelihood with Pólya–Gamma auxiliary variables to obtain fully conjugate Gibbs updates for β and a β-marginalized collapsed slice update for ρ (the |I − ρW| Jacobian cancels, so ρ is Krylov-accelerable).

The sampler bypasses PyMC’s NUTS entirely. It produces an xarray.DataTree object compatible with all downstream diagnostics (spatial_diagnostics(), impacts(), summary()).

With the Gibbs sampler, fit() does not accept target_accept, a NUTS-only argument; passing it raises TypeError.

The PG shape parameter is always h = 1 (one trial per observation).

__init__(formula=None, data=None, y=None, X=None, W=None, *, unit_effects=None, time_effects=False, dynamic=False, latent=None, **kwargs)[source]

Methods

__init__([formula, data, y, X, W, ...])

fit([draws, tune, chains, target_accept, ...])

Draw samples from the posterior.

fitted_probabilities(*args, **kwargs)

Fitted P(y = 1) at the posterior-mean parameters.

fitted_values()

Return fitted values at posterior mean parameters.

impacts([return_posterior_samples, scale, ...])

Compute Bayesian inference for direct, indirect, and total impacts.

residuals()

Return residuals y - fitted_values.

spatial_diagnostics()

Run Bayesian LM specification tests and return a summary table.

spatial_diagnostics_decision([alpha, format])

Return a model-selection decision from Bayesian LM test results.

summary([var_names])

Return posterior summary table.

Attributes

inference_data

Return the ArviZ DataTree from the most recent fit.

pymc_model

Return the PyMC model object built for the most recent fit.

fit(draws=2000, tune=1000, chains=4, target_accept=None, random_seed=None, progressbar=True, sampler=None, backend=None, thin=1, n_jobs=-1, idata_kwargs=None, **sample_kwargs)[source]

Draw samples from the posterior.

Dispatches to this model’s Gibbs sampler (sampler="gibbs") or NUTS (sampler="nuts"). When sampler is None (default), Gibbs is used if the model has a registered Gibbs sampler, otherwise NUTS.

Parameters:
draws : int

Post-warmup draws, warmup steps, and number of chains.

tune : int

Post-warmup draws, warmup steps, and number of chains.

chains : int

Post-warmup draws, warmup steps, and number of chains.

target_accept : float, optional

Target acceptance rate for NUTS. NUTS-only: passing it with the Gibbs sampler raises TypeError. Defaults to 0.9 for NUTS.

random_seed : int, optional

Seed for reproducibility.

progressbar : bool, default True

Show progress bar(s) during sampling.

sampler : {"gibbs", "nuts", None}, default None

Sampling method. None auto-selects Gibbs when this model has one, else NUTS.

backend : str, optional

What executes the sampler. For Gibbs, "auto" (the default), "jax" or "numpy"; "auto" takes JAX when it is installed and supported by the family, else NumPy. For NUTS, "pymc" (the default), "nutpie", "numpyro" or "blackjax".

thin : int, default 1

Keep every thin-th post-warmup Gibbs draw (Gibbs only).

n_jobs : int, default -1

Parallel workers for the NumPy Gibbs path (Gibbs only).

idata_kwargs : dict, optional

{"log_likelihood": True} stores the complete Jacobian-corrected pointwise log-likelihood that az.loo / az.compare need, for Gibbs and NUTS alike. Off by default, as in PyMC: it holds one value per draw, chain, and observation (16 GB at n = 250,000 with 4 × 2,000 draws). For NUTS the dict is also passed to pm.sample.

**sample_kwargs

For NUTS, forwarded to pm.sample; for Gibbs, the family’s declared options (an unsupported key raises).

Return type:

xarray.DataTree

fitted_probabilities(*args, **kwargs)[source]

Fitted P(y = 1) at the posterior-mean parameters.

fitted_values()[source]

Return fitted values at posterior mean parameters.

Returns:

Posterior-mean fitted values (on the model’s native scale; fixed-effects-transformed for panel models).

Return type:

np.ndarray

impacts(return_posterior_samples=False, scale='logodds', method='auto')[source]

Compute Bayesian inference for direct, indirect, and total impacts.

Parameters:
return_posterior_samples : bool, optional

If True, also return the posterior draws for each effect type.

scale : {"logodds", "probability"}, default "logodds"

Scale on which impacts are reported.

"logodds" returns impacts on the linear-predictor (log-odds) scale \(\eta = (I - \rho W)^{-1} X\beta\). This is the cheap default; effects are linear in \(\beta\) and do not require evaluating the link.

"probability" returns response-scale impacts

\[\partial p / \partial x_r = \operatorname{diag}(p(1-p))\,(I - \rho W)^{-1}\beta_r,\]

which match the LeSage-Pace (2009) and PG-paper formulas. This is more expensive because it requires the diagonal of the spatial multiplier and the fitted probabilities for each draw.

method : {"auto", "eigen", "sparse"}, default "auto"

Only used when scale="probability". "eigen" uses the shared eigendecomposition cache (O(n²) per draw); "sparse" uses one sparse LU per draw plus a Hutchinson diagonal estimator (O(nnz) per draw); "auto" picks sparse when \(n\) exceeds _PROBABILITY_EFFECTS_EIGEN_MAX_N (default 2000).

property inference_data[source]

Return the ArviZ DataTree from the most recent fit.

Returns:

The inference data object, or None if the model has not been fit yet.

Return type:

xarray.DataTree or None

property pymc_model[source]

Return the PyMC model object built for the most recent fit.

For a Gibbs-fitted cross-section the PyMC model is not constructed during sampling; it is built lazily on first access so that downstream consumers (e.g. bridge sampling for marginal likelihoods) can evaluate logp and the prior under the same model definition used by the NUTS path.

Returns:

The model object used by fit(), or None if the instance has not been fit yet.

Return type:

pymc.Model or None

residuals()[source]

Return residuals y - fitted_values.

Returns:

Residual vector y - fitted_values on the same scale as fitted_values().

Return type:

np.ndarray

spatial_diagnostics()[source]

Run Bayesian LM specification tests and return a summary table.

Looks up the diagnostic suite registered for this model class and calls each test function on this fitted model, collecting the results into a tidy DataFrame. The set of tests depends on the model type — for example, an OLS model runs LM-Lag, LM-Error, LM-SDM-Joint, and LM-SLX-Error-Joint, while an SAR model runs LM-Error, LM-WX, and Robust-LM-WX. Panel models run the Panel--prefixed analogues (e.g. Panel-LM-Lag).

Requires the model to have been fit (.fit() called) and a spatial weights matrix W to have been supplied at construction time.

Returns:

DataFrame indexed by test name with columns:

Column

Description

statistic

Posterior mean of the LM statistic

median

Posterior median of the LM statistic

df

Degrees of freedom for the \(\chi^2\) reference

p_value

Bayesian p-value: 1 - chi2.cdf(mean, df)

ci_lower

Lower bound of 95% credible interval (2.5%)

ci_upper

Upper bound of 95% credible interval (97.5%)

The DataFrame has attrs["model_type"] (class name) and attrs["n_draws"] (total posterior draws) metadata.

Return type:

pandas.DataFrame

Raises:
  • RuntimeError – If the model has not been fit yet.

  • ValueError – If no spatial weights matrix W was supplied.

See also

spatial_diagnostics_decision

Model-selection decision based on the test results.

impacts

Posterior inference for direct/indirect/total impacts.

Examples

>>> ols = OLS(formula="price ~ income + crime", data=df, W=w)
>>> ols.fit()
>>> ols.spatial_diagnostics()
                 statistic  median  df  p_value  ci_lower  ci_upper
LM-Lag                3.21    2.98   1    0.073      0.12      8.54
LM-Error              5.67    5.34   1    0.017      0.34     12.10
LM-SDM-Joint          7.89    7.12   4    0.096      1.23     18.32
LM-SLX-Error-Joint    6.45    5.98   4    0.168      0.89     15.67
spatial_diagnostics_decision(alpha=0.05, format='graphviz')[source]

Return a model-selection decision from Bayesian LM test results.

One rule reads every fitted model’s tests: the specific-to-general rule of Anselin et al. [1996], which spreg’s stge_classic implements. A model that leaves two directions untested runs a plain and a robust LM test for each:

  1. Neither plain test is significant → keep the model.

  2. One is → add that direction.

  3. Both are → add the direction whose robust test is significant; if both or neither are, the one with the smaller robust p-value.

A plain test is valid when the other direction is absent and a robust one is valid either way, at a cost in power, so the robust pair only chooses between two plain rejections. The SDM and SDEM leave one direction, and its test decides. The directions, and the models they lead to:

  • OLS: lag → SAR, error → SEM, WX → SLX. With three directions untested, the lag and the error together go to their robust pair whether or not WX also rejects (the SAR or SEM then tests it); the lag and WX go to the robust pair of Koley and Bera [2024], whose lag test is LM-Error at the OLS fit; the error and WX, whose scores are uncorrelated there, to the smaller p-value.

  • SAR: WX → SDM, error → SARAR.

  • SEM: lag → SARAR, WX → SDEM.

  • SLX: lag → SDM, error → SDEM.

  • SDM: error → GNS. SDEM: lag → GNS.

SARAR and GNS are where the tests point; neighbayes does not fit them. The WX tests are those of Koley and Bera [2024]. Panel models run the Panel- tests through the same trees ([Elhorst, 2014]), and the Tobit classes the same trees with Tobit leaves. The logit, probit and NB at their aspatial null read their tests as a pair of their own.

Parameters:
alpha : float, default 0.05

Significance level for the Bayesian p-values.

format : {"graphviz", "ascii", "model"}, default "graphviz"

Output format. "model" returns the recommended-model name string. "ascii" returns an indented box-drawing rendering of the full decision tree with the chosen path highlighted. "graphviz" returns a graphviz.Digraph object that renders inline in Jupyter; if the optional graphviz package is not installed a UserWarning is issued and the ASCII rendering is returned instead.

Returns:

Recommended model name when format="model", an ASCII tree string when format="ascii", or a graphviz.Digraph when format="graphviz" (with ASCII fallback on missing dep).

Return type:

str or graphviz.Digraph

See also

spatial_diagnostics

Compute the Bayesian LM test statistics.

References

Koley and Bera [2024], Anselin et al. [1996], Elhorst [2014]

summary(var_names=None, **kwargs)[source]

Return posterior summary table.

Parameters:
var_names : list, optional

Variable names to include in the summary.

**kwargs

Additional arguments passed to arviz.summary().

Returns:

Posterior summary statistics.

Return type:

pandas.DataFrame